{smcl}
{* 9 July 2008}{...}
{cmd:help cdfplot}
{hline}

{title:Title}

  {hi:Plots the cumulative distribution function}

{title:Syntax}

{p 8 27 2}
{cmdab:cdfplot} {help varname} [{help if}] [{help in}] [{help weight}] 
[{cmd:,} {it:options}]

{synoptset 20 tabbed}{...}
{synopthdr}
{synoptline}
{syntab:Main}
{synopt:{opt by}(varname)} specifies a separate c.d.f. to be drawn for each value of varname. {p_end}
{synopt:{opt norm:al}} specifies that a Gaussian probability curve with the same mean and standard 
deviation to be superimposed over the c.d.f. {p_end}
{synopt:{opt same:sd}} specifies that the normal c.d.f.s use the same same standard deviation. {p_end}
{synopt:{opt nocdf}} specifies that the empirical c.d.f. is not drawn. {p_end}
{synopt:{opt opt1}(twoway_options)} specifies the additional graph options for the c.d.f. plot. {p_end}
{synopt:{opt opt2}(twoway_options)} specifies the additional graph options for the Gaussian c.d.f. plots . {p_end}
{synopt:{opt {help twoway_options}}} any twoway options are applied to the final graph. {p_end}
{synoptline}
{p2colreset}{...}


{title:Description}

{pstd}
{hi:cdfplot} plots the sample cumulative distribution function. Distributions can
be compared within subgroups defined by a second variable. The best fitting 
normal (Gaussian) model may be superimposed over the sample c.d.f.

{title:Updating this command using SSC}

{pstd} 
To obtain the latest version click the following to uninstall the old version

{pstd}
{stata ssc uninstall cdfplot}

{pstd}
And click here to install the new version

{pstd}
{stata ssc install cdfplot}

{title:Options}

{phang}
{opt by}(varname) specifies a separate c.d.f. to be drawn for each value 
of varname. These are plotted on the same graph for easier comparison.

{phang}
{opt norm:al} specifies that a normal probability curve with the same mean and standard 
deviation to be superimposed over the c.d.f.

{phang}
{opt same:sd} is relevant only when {hi:by} and {hi:normal} options are used together.
It specifies that the normal curves with different means have the same standard deviations. 
This demonstrates the fit of the conventional Gaussian location shift model.

{phang}
{opt nocdf} specifies that the empirical c.d.f. is not drawn but this option will not work unless
the normal option is specified, hence only the smoothed Gaussian c.d.f will be drawn. {p_end}

{phang}
{cmdab:opt1}{cmd:(}{it:twoway_options}{cmd:)} specifies additional graph options for the c.d.f. plots.

{phang}
{cmdab:opt2}{cmd:(}{it:twoway_options}{cmd:)} specifies additional graph options for the Gaussian c.d.f. plots.

{phang}
If the {hi:xlog} option is used, the {hi:normal} option causes log-normal distributions
to be fitted.

{title:Examples}

{pstd}
Using the variable {hi:length} from the {hi:auto} data (click on the following commands in order):

{pstd}
{hi: NOTE DATA will be lost when loading new data}

{pstd}
{stata sysuse auto,replace}{p_end}
{pstd}
{stata cdfplot length, normal}{p_end}
{pstd}
{stata cdfplot length, by(foreign)}{p_end}
{pstd}
{stata cdfplot length, by(foreign) norm saving(mygraph)}{p_end}
{pstd}
{stata cdfplot length [fw=rep78], by(foreign) norm saving(mygraph,replace)}{p_end}

{pstd}
Using the variable {hi:bp} from the {hi:bplong} data 

{pstd}
{stata sysuse bplong,replace}{p_end}
{pstd}
{stata cdfplot bp, norm}{p_end}
{pstd}
{stata cdfplot bp, by(agegrp)}{p_end}
{pstd}
{stata cdfplot bp, by(agegrp) norm}{p_end}

{pstd}
Now to show the use of the options in the individual graphs, {hi:opt1} alters the line colours
for the c.d.f. graphs and {hi:opt2} alters the line pattern for the Gaussian c.d.f.

{pstd}
{stata cdfplot bp, by(agegrp) opt1( lc(red green olive) ) opt2( lp(dash dash dash) ) norm }

{title:Author}

{pstd}
Adrian Mander, MRC Human Nutrition Research Unit, Cambridge, UK.

{pstd}
Email {browse "mailto:adrian.mander@mrc-hnr.cam.ac.uk":adrian.mander@mrc-hnr.cam.ac.uk}


{title:Acknowledgement}

{pstd}
This command is nearly a direct port of the {hi:cdf} command that was written by
{bf: David Clayton} and {bf:Michael Hills} in STB-49.

{title:Also see}

{pstd}
{help cdf} (if installed)

